serverless ops Ebook free download pdf pdf
O’Reilly Web Ops
Serverless Ops
A Beginner’s Guide to AWS Lambda and Beyond
Michael Hausenblas
Serverless Ops
by Michael Hausenblas Copyright © 2017 O’Reilly Media, Inc. All rights reserved. Printed in the United States of America. Published by O’Reilly Media, Inc., 1005 Gravenstein Highway North, Sebastopol, CA 95472. O’Reilly books may be purchased for educational, business, or sales promotional use. Online editions are also available for most titles ( ). For more information, contact our corporate/institutional sales department: 800-998-9938 or corporate@oreilly.com.
Editor: Virginia Wilson Acquisitions Editor: Brian Anderson Production Editor: Shiny Kalapurakkel Copyeditor: Amanda Kersey Proofreader: Rachel Head Interior Designer: David Futato Cover Designer: Karen Montgomery Illustrator: Rebecca Panzer
November 2016: First Edition
Revision History for the First Edition
2016-11-09: First Release The O’Reilly logo is a registered trademark of O’Reilly Media, Inc. Serverless Ops, the cover image, and related trade dress are trademarks of O’Reilly Media, Inc.
While the publisher and the author have used good faith efforts to ensure that the information and instructions contained in this work are accurate, the publisher and the author disclaim all responsibility for errors or omissions, including without limitation responsibility for damages resulting from the use of or reliance on this work. Use of the information and instructions contained in this work is at your own risk. If any code samples or other technology this work contains or describes is subject to open source licenses or the intellectual property rights of others, it is your responsibility to ensure that your use thereof complies with such licenses and/or rights.
[LSI]
Preface
The dominant way we deployed and ran applications over the past decade was machine-centric. First, we provisioned physical machines and installed our software on them. Then, to address the low utilization and accelerate the roll-out process, came the age of virtualization. With the emergence of the public cloud, the offerings became more diverse: Infrastructure as a Service (IaaS), again machine-centric; Platform as a Service (PaaS), the first attempt to escape the machine-centric paradigm; and Software as a Service (SaaS), the so far (commercially) most successful offering, operating on a high level of abstraction but offering little control over what is going on.
Over the past couple of years we’ve also encountered some developments that changed the way we think about running applications and infrastructure as such: the microservices architecture, leading to small-scoped and loosely coupled distributed systems; and the world of containers, providing application-level dependency management in either on-premises or cloud environments.
With the advent of DevOps thinking in the form of Michael T. Nygard’s (Pragmatic Programmers) and the , we’ve witnessed the transition to immutable infrastructure and the need for organizations to encourage and enable developers and ops folks to work much more closely together, in an automated fashion and with mutual understanding of the motivations and incentives.
In 2016 we started to see the serverless paradigm . Starting with the AWS Lambda announcement in 2014, every major cloud player has now introduced such offerings, in addition to many new players like OpenLambda or Galactic Fog specializing in this space. Before we dive in, one comment and disclaimer on the term “serverless” itself: catchy as it is, the name is admittedly a misnomer and has attracted a fair amount of criticism, including from people such as AWS CTOt is as misleading as “NoSQL” because it defines the concept in
terms of what it is not about. There have been a number of attempts to rename it; for example, to
(FaaS). Unfortunately, it seems we’re stuck with the term because it has gained
traction, and the majority of people interested in the paradigm don’t seem to have a problem with it.You and Me
My hope is that this report will be useful for people who are interested in going serverless, people who’ve just started doing serverless computing, and people who have some experience and are seeking guidance on how to get the maximum value out of it. Notably, the report targets:
DevOps folks who are exploring serverless computing and want to get a quick overview of the space and its options, and more specifically novice developers and operators of AWS Lambda environments or want to apply the paradigm in a new project This report aims to provide an overview of and introduction to the serverless paradigm, along with best-practice recommendations, rather than concrete implementation details for offerings (other than exemplary cases). I assume that you have a basic familiarity with operations concepts (such as deployment strategies, monitoring, and logging), as well as general knowledge about public cloud offerings. Note that true coverage of serverless operations would require a book with many more pages. As such, we will be covering mostly techniques related to AWS Lambda to satisfy curiosity about this emerging technology and provide useful patterns for the infrastructure team that administers these architectures.
As for my background: I’m a developer advocate at Mesosphere working on distributed operating system for both containerized workloads and elastic data pipelines. I started to dive into serverless offerings in early 2015, doing proofs of concepts, bout the topic, as well as helping with the onboarding of serverless offerings onto DC/OS.
Acknowledgments
I’d like to thank Charity Majors for sharing her insights around operations, DevOps, and how developers can get better at operations. Her talks and articles have shaped my understanding of both the technical and organizational aspects of the operations space. The technical reviewers of this report deserve special thanks too. Eric Windisch (IOpipe, Inc.), Aleksander Slominski (IBM), and Brad Futch (Galactic Fog) haven taken out time of their busy schedules to provide very valuable feedback and certainly shaped it a lot. I owe you all big time (next Velocity conference?). A number of good folks have supplied me with examples and references and have written timely articles that served as brain food: to Bridget Kromhout, Paul Johnston, and Rotem Tamir, thank you so much for all your input. A big thank you to the O’Reilly folks who looked after me, providing guidance and managing the process so smoothly: Virginia Wilson and Brian Anderson, you rock! Last but certainly not least, my deepest gratitude to my awesome family: our sunshine artist Saphira, our sporty girl Ranya, our son Iannis aka “the Magic rower,” and my ever-supportive wife Anneliese. Couldn’t have done this without you, and the cottage is my second-favorite place when I’m at home. ;)
1 The term NoSQL suggests it’s somewhat anti-SQL, but it’s not about the SQL language itself.
Instead, it’s about the fact that relational databases didn’t use to do auto-sharding and hence were not easy or able to be used out of the box in a distributed setting (that is, in cluster mode).
Chapter 1. Overview
Before we get into the inner workings and challenges of serverless computing, or Function as a Service (FaaS), we will first have a look at where it sits in the spectrum of computing paradigms, comparing it with traditional three-tier apps, microservices, and Platform as a Service (PaaS) solutions. We then turn our attention to the concept of serverless computing; that is, dynamically allocated resources for event-driven function execution.
A Spectrum of Computing Paradigms
The basic idea behind serverless computing is to make the unit of computation a function. This effectively provides you with a lightweight and dynamically scalable computing environment with a certain degree of control. What do I mean by this? To start, let’s have a look at the spectrum of computing paradigms and some examples in each area, as depicted in .
Figure 1-1. A spectrum of compute paradigms
In a monolithic application, the unit of computation is usually a machine (bare-metal or virtual). With microservices we often find containerization, shifting the focus to a more fine-grained but still machine-centric unit of computing. A PaaS offers an environment that includes a collection of APIs and objects (such as job control or storage), essentially eliminating the machine from the picture. The serverless paradigm takes that a step further: the unit of computation is now a single function whose lifecycle you manage, combining many of these functions to build an application.
Looking at some (from an ops perspective), relevant dimensions further sheds light on what the different paradigms bring to the table: Agility measured in months; serverless environments allow much more rapid deployments. Control
With the machine-centric paradigms, you have a great level of control over the environment. You can set up the machines to your liking, providing exactly what you need for your workload (think libraries, security patches, and networking setup). On the other hand, PaaS and serverless solutions offer little control: the service provider decides how things are set up. The flip side of control is maintenance: with serverless implementations, you essentially outsource the maintenance efforts to the service provider, while with machine-centric approaches the onus is on you. In addition, since autoscaling of functions is typically supported, you have to do less engineering yourself.
Cost per unit For many folks, this might be the most attractive aspect of serverless offerings—you only pay for the actual computation. Gone are the days of provisioning for peak load only to experience low resource utilization most of the time. Further, A/B testing is trivial, since you can easily deploy multiple versions of a function without paying the overhead of unused resources.
The Concept of Serverless Computing
With this high-level introduction to serverless computing in the context of the computing paradigms out of the way, we now move on to its core tenents. At its core, serverless computing is event-driven, as shown in .
In general, the main components and actors you will find in serverless offerings are: Management interfaces
Register, upgrade, and control functions via web UIs, command-line interfaces, or HTTP APIs. Triggers
Define when a function is invoked, usually through (external) events, and are scheduled to be executed at a specific time. Integration points Support control and data transfer from function-external systems such as storage.
So, the serverless paradigm boils down to reacting to events by executing code that has been uploaded and configured beforehand.
HOW SERVERLESS IS DIFFERENT FROM PAAS
Quite often, when people start to dig into serverless computing, I hear questions like “How is this different from PaaS?” Serverless computing (or FaaS), refers to the idea of dynamically allocating resources for an event-driven function execution. A number of related paradigms and technologies exist that you may have come across already. This sidebar aims to compare and delimit them. PaaS shares a lot with the serverless paradigm, such as no provisioning of machines and autoscaling. However, the unit of computation is much smaller in the latter. Serverless computing is also job-oriented rather than application-oriented. For more on this topic, see Carl Osipov’s
The Remote Procedure Call (RPC) protocol is all about the illusion that one can call a remotely executed function (potentially on a different machine) in the same way as a locally executed function (in the same memory space). Stored procedures have things in common with serverless computing (including some of the drawbacks, such as lock-in), but they’re database-specific and not a general-purpose computing paradigm. Microservices are not a technology but an architecture and can, among other things, be implemented with serverless offerings.
Containers are typically the basic building blocks used by serverless offering providers to enable rapid provisioning and isolation.
Conclusion
In this chapter we have introduced serverless computing as an event-driven function execution paradigm with its three main components: the triggers that define when a function is executed, the management interfaces that register and configure functions, and integration points that interact with external systems (especially storage). Now we’ll take a deeper look at the concrete offerings in this space.
1 I’ve deliberately left routing (mapping, for example, an HTTP API to events) out of the core tenents since different offerings have different approaches for how to achieve this.
Chapter 2. The Ecosystem
In this chapter we will explore the current serverless computing offerings and the wider ecosystem. We’ll also try to determine whether serverless computing only makes sense in the context of a public cloud setting or if operating and/or rolling out a serverless offering on-premises also makes sense.
Overview
Many of the serverless offerings at the time of writing of this report (mid-2016) are rather new, and the space is growing quickly.
gives a brief comparison of the main players. More detailed breakdowns are provided in the following sections.
Table 2-1. Serverless offerings by company Offering Cloud offering On-premises Launched Environments AWS Lambda Yes No 2014 Node.js, Python, Java Azure Functions Yes Yes 2016 C#, Node.js, Python, F#, PHP, Java Google Cloud Functions Yes No 2016 JavaScript iron.io No Yes 2012 Ruby, PHP, Python, Java, Node.js, Go, .NET Galactic Fog’s Gestalt No Yes 2016 Java, Scala, JavaScript, .NET
IBM OpenWhisk Yes Yes 2014 Node.js, Swift
Note that by cloud offering, I mean that there’s a managed offering in one of the public clouds available, typically with a pay-as-you-go model attached.
AWS Lambda
Introduced in 2014 in an WS Lambda is the incumbent in the serverless space and makes up an ecosystem in its own right, including frameworks and tooling on top of it, built by folks outside of Amazon. Interestingly, the motivation to introduce Lambda originated in observations of EC2 usage: the AWS team noticed that increasingly event-driven workloads were being deployed, such as infrastructure tasks (log analytics) or batch processing jobs (image manipulation and the like). AWS Lambda started out with support for the Node runtime and currently
The main building blocks of AWS Lambda are: The AWS Lambda Web UI (see itself to register, execute, and manage functions
, including, but not limited to, events from S3, SNS, and CloudFormation to trigger
the execution of a function CloudWatch for logging and monitoring
Figure 2-1. AWS Lambda dashboard
PricingAWS Lambda is based on the total number of requests as well as execution time. The first
1 million requests per month are free; after that, it’s $0.20 per 1 million requests. In addition, the free tier includes 400,000 GB-seconds of computation time per month. The minimal duration you’ll be billed for is 100 ms, and the actual costs are determined by the amount of RAM you allocate to your function (with a minimum of 128 MB).
Availability Lambda has been available since 2014 and is a public cloud–only offering.
We will have a closer look at the AWS Lambda offering in
Chapter 4 , where we will walk through an example from end to end.
During the Build 2016 conference Microsoft release , supporting functions written with C#, Node.js, Python, F#, PHP, batch, bash, Java, or any executable. The Functions runtime is
and integrates with Azure-internal and -external services such as Azure Event Hubs,
Azure Service Bus, Azure Storage, and GitHub webhooks. The Azure Functions portal, depicted in , comes with templates and monitoring capabilities.
Figure 2-2. Azure Functions portal
As an aside, Microsoft also offers other serverless solutions such as Azure Web Jobs and Microsoft Flow (an “if this, then that” [IFTTT] for business competitors).
Pricing
Azure Functions is similar to that of AWS Lambda; you pay based on code execution time
and number of executions, at a rate of $0.000008 per GB-second and $0.20 per 1 million executions.As with Lambda, the free tier includes 400,000 GB-seconds and 1 million executions.
Availability
Since early 2016, the Azure Functions service has been available both as a public cloud offering and on-premises as part of the
Google Cloud Functions
an be triggered by messages on a Cloud Pub/Sub topic or through mutation
events on a Cloud Storage bucket (such as “bucket is created”). For now, the service only supports
Functions directly from your GitHub or Bitbucket repository without needing to upload code or manage versions yourself. Logs emitted are automatically written to Stackdriver Logging and performance telemetry is recorded in Stackdriver Monitoring.
shows the Google Cloud Functions view in the Google Cloud console. Here you can create a function, including defining a trigger and source code handling.
Figure 2-3. Google Cloud Functions
PricingSince the Google Cloud Functions service is in Alpha, no pricing has been disclosed yet. However, we can assume that it will be priced competitively with the incumbent, AWS Lambda.
Availability
Google introduced Cloud Functions in February 2016. At the time of writing, it’s in Alpha status with access on a per-request basis and is a public cloud–only offering.
Iron.io as supported serverless concepts and frameworks since 2012. Some of the early offerings,
such as IronQueue, IronWorker, and IronCache, encouraged developers to bring their code and run it in the Iron.io-managed platform hosted in the public cloud. Written in Go, Iron.io recently embraced Docker and integrated the existing services to offer a cohesive microservices platform. Codenamed Project Kratos, the serverless computing framework from Iron.io aims to bring AWS Lambda to
In , the overall Iron.io architecture is depicted: notice the use of containers and container images.
Figure 2-4. Iron.io architecture Pricing
No public plans are available, but you can use the offering via a number of deployment options, including
Availability
Iron.io has offered its services since 2012, with a recent update around containers and supported environments.
Galactic Fog’s Gestalt ) is a serverless offering that bundles containers with security and data features, allowing developers to write and deploy microservices on-premises or in the cloud.
Figure 2-5. Gestalt Lambda
Pricing No public plans are available. AvailabilityLaunched in mid-2016, the Gestalt Framework isnd is suitable for cloud and on-premises deployments; no hosted service is available yet. See the MesosCon 2016 talk by Brad Futch for details on the current state as well as the upcoming rewrite of Gestalt Lambda called LASER.
IBM OpenWhisk
IBM s an open source alternative to AWS Lambda. As well as supporting Node.js, OpenWhisk can run snippets written in Swift. You can install it on your local machine running Ubuntu. The service is integrated with IBM Bluemix, the PaaS environment powered by Cloud Foundry. Apart from invoking Bluemix services, the framework can be integrated with any third-party service that supports webhooks. Developers can use a CLI to target the OpenWhisk framework.
shows the high-level architecture of OpenWhisk, including the trigger, management, and integration point options.
Figure 2-6. OpenWhisk architecture
PricingThe based on Bluemix, at a rate of $0.0288 per GB-hour of RAM and $2.06 per public IP address. The free tier includes 365 GB-hours of RAM, 2 public IP addresses, and 20 GB of external storage.
Availability
Since 2014, OpenWhisk has been available as a and for on-premises deployments with Bluemix as a dependency. See for more details on the offering.
Other Players
In the past few years, the serverless space has seen quite some uptake, not only in terms of end users but also in terms of providers. Some of the new offerings are open source, some leverage or extend existing offerings, and some are specialized offerings from existing providers. They include:
, an automated computing service that runs and scales your microservices Auth0, a serverless environment supporting Node.js with a focus on security
(IoT) applications powered by AWS Lambda and AWS API Gateway, with plans to support other providers, such as Azure and Google Cloud
n analytics and distributed tracing service that allows you to see inside AWS Lambda
functions for better insights into the daily operations
Cloud or on-Premises?
A question that often arises is whether serverless computing only makes sense in the context of a public cloud setting, or if rolling out a serverless offering on-premises also makes sense. To answer this question, we will discuss elasticity features, as well as dependencies introduced when using a serverless offering.
So, which one is the better option? A public cloud offering such as AWS Lambda, or one of the existing open source projects, or your home-grown solution on-premises? As with any IT question, the answer depends on many things, but let’s have a look at a number of considerations that have been brought up in the community and may be deciding factors for you and your organization.
One big factor that speaks for using one of the (commercial) public cloud offerings is the ecosystem. Look at the triggers) as well as the integrations with other services, such as S3, Azure SQL Database, and monitoring and security features. Given that the serverless offering is just one tool in your toolbelt, and you might already be using one or more offerings from a certain cloud provider, the ecosystem is an important point to consider.
Oftentimes the argument is put forward that true autoscaling of the functions only applies to public cloud offerings. While this is not black and white, there is a certain point to this claim: the elasticity of the underlying IaaS offerings of public cloud providers will likely outperform whatever you can achieve in your datacenter. This is, however, mainly relevant for very spiky or unpredictable workloads, since you can certainly add virtual machines (VMs) in an on-premises setup in a reasonable amount of time, especially when you know in advance that you’ll need them.
Avoiding lock-in is probably the strongest argument against public cloud serverless deployments, not so much in terms of the actual code (migrating this from one provider to another is a rather straightforward process) but more in terms of the triggers and integration points. At the time of writing, there is no good abstraction that allows you to ignore storage or databases and work around triggers that are available in one offering but not another. Another consideration is that when you deploy the serverless infrastructure in your datacenter you have full control over, for example how long a function can execute. The public cloud offerings at the current point in time do not disclose details about the underlying implementation, resulting in a lot of guesswork and trial and error when it comes to optimizing the operation. With an on-premises deployment you can go as far as developing your own solution, as discussed in however, you should be aware of the investment (both in terms of development and operations) that is
summarizes the criteria discussed in the previous paragraphs.
Offering Cloud On-premises Ecosystem Yes No True autoscaling Yes No Avoiding lock-in No Yes End-to-end control No Yes
Note that depending on what is important to your use case, you’ll rank different aspects higher or lower; my intention here is not to categorize these features as positive or negative but simply to point out potential criteria you might want to consider when making a decision.
Conclusion
In this chapter, we looked at the current state of the serverless ecosystem, from the incumbent AWS Lambda to emerging open source projects such as OpenLambda. Further, we discussed the topic of using a serverless offering in the public cloud versus operating (and potentially developing) one on- premises based on decision criteria such as elasticity and integrations with other services such as databases. Next we will discuss serverless computing from an operations perspective and explore how the traditional roles and responsibilities change when applying the serverless paradigm.
Chapter 3. Serverless from an Operations
Perspective
The serverless paradigm blurs the line between development and operations. On the one hand, certain traditionally necessary steps such as provisioning a machine do not apply anymore; on the other hand, developers can’t simply hand off binaries to operations. In this chapter, we will first discuss roles in the context of a serverless setup and then have a closer look at typical activities, good practices, and antipatterns around serverless ops.
AppOps
With serverless computing, it pays off to rethink roles and responsibilities in the team. To do that, I’m borrowing a term that was first coined by of Digital Ocean: AppOps. The basic idea behind AppOps is that the one who writes a service also operates it in production. This means that AppOps are for the services they have developed. In order for this to work, the infrastructure used needs to support service- or app-level monitoring of metrics as well as alerting if the service doesn’t perform as expected. Further, there’s another role necessary: a group of people called the infrastructure team. This team manages the overall infrastructure, owns global policies, and advises the AppOps.
A sometimes-used alternative label for the serverless paradigm is “NoOps,” suggesting that since there are no machines to provision, the need for operations folks is not given. This term is, however, misleading and best avoided. As discussed, operational skills and practices are not only necessary but pivotal in the serverless context—just not in the traditional sense.
Operations: What’s Required and What Isn’t
To define operations in the serverless context, I’ll start out with definition:
Operations is the constellation of your org’s technical skills, practices and cultural values around designing, building and maintaining systems, shipping software , and solving problems with technology.
—Serverlessness, NoOps and the Tooth Fairy,, May 2016 Building on this definition, we can now understand what is required for successful operations: Scalability autoscaling support usually found in serverless offerings should not be taken as an excuse to not study and understand this property.
Resilience Having a good understanding of the failure modes and self-healing methods. As with scaling, a lot of this is taken care of by the serverless offering; however, one needs to know the limitations of this.
Availability Another area where in a serverless setup the control points are limited. The current offerings come with few service-level objectives or agreements, and status pages are typically not provided. The monitoring focus should hence be more on the platform than on the function level.
Maintainability Of the function code itself. Since the code is very specific and has a sharp focus, the length of the function shouldn’t be a problem. However, understanding how a bunch of functions work together to achieve some goal is vital.
Visibility Typically limited by what the serverless provider allows; very often little is known about the underlying infrastructure (OS level, container, etc.).
Interestingly, the way serverless computing addresses many of these aspects seems to be what makes it most attractive. The result of a Twitter poll carried out by DevOps legend in May 2016 highlights this (see .
Figure 3-1. Twitter poll: What makes serverless different for you?
As pointed out by here are a number of responsibilities found in traditional admin roles that are not applicable in a serverless setup: OS-level configuration management and (security) patches are not required, since the execution environment is fixed and managed by the serverless provider. Backups are not necessary since functions are supposed to be stateless. Service-level scaling is typically a feature of the serverless platform. Many activities that were traditionally expected to be carried out by the operations team, such as deployments or monitoring, are now the responsibility of the AppOps. However, the infrastructure team has a number of new responsibilities that we will discuss in the next section.
Infrastructure Team Checklist
As a member of the infrastructure team, you act as a coach and guide to AppOps. Here are a couple of
Make sure that the functions are versioned properly. A function’s source code should reside in a (ideally distributed) version control system such as Git. This is an infrastructure task that you should manage, along with enforcing the respective policies around access and push rights.
Keep track of the overall picture—that is, the full set of functions, potentially owned by a number of AppOps—so you can provide recommendations about when to go serverless (as described in
Chapter 4 ) and when it makes more (economic) sense to move back to a dedicated-machine solution. Support the troubleshooting process. Since serverless functions typically depend on external
systems such as (managed) storage, you can help establishurther, there may be cases where you can provide insights—for example, in the form of access to additional logs—when an AppOp debugs a function that is either not working correctly or has a higher than normal execution error rate.
Provide insights regarding serverless functions. The infrastructure team’s holistic view is particularly valuable here. Identify potential cost optimizations. While with serverless solutions, there’s no capacity planning in the traditional sense, AppOps can make better-informed decisions about the few resource consumption parameters (such as RAM) under their control when the infrastructure team can offer guidance in terms of overall usage.
Conclusion
In this chapter we had a look at the new roles encouraged and to a certain extent required by the serverless paradigm. The traditional developer role morphs into an AppOps role, responsible for not only writing the code but also monitoring and troubleshooting it. In addition, the infrastructure team doesn’t have to perform certain tasks required in, say, VM-based deployments, such as patching or scaling, and therefore can take on new responsibilities such as load testing and act as advisors for AppOps. Now we’re in a position to look at application areas where serverless computing is a good fit and what the limitations and challenges of this new paradigm are.
Chapter 4. Serverless Operations Field
Guide
This chapter is meant as a guide to help you decide when and where to use serverless computing. We will talk about application areas and review concrete use cases for it. Then we’ll turn our attention to the limitations of serverless computing, potential gotchas, and a migration guide from a monolithic application. Last but not least, we will have a look at a simple walkthrough example to discuss the implications for operations as outlined in the previous chapter.
Latency Versus Access Frequency
Before you embark on the serverless journey, you might want to ask yourself how applicable the serverless paradigm is for the use case at hand. There may be an array of deciding factors for your use case, which can be summed up in two categories: technical and economic. Technical requirements could be supported programming languages, available triggers, or integration points supported by a certain offering. On the other hand, you or the budget holder are probably also interested in the costs of using the service (at least in the context of a public cloud offering, where these are often more ).
provides a rough guide for the applicability of serverless computing along two dimensions: latency and access frequency.
Figure 4-1. Latency sensitivity versus access frequency
By latency, I mean how much time can acceptably elapse between function invocation and termination. It might be important for your use case that you have guarantees around latency—for example, that the 90th percentile cannot exceed 100 ms. It might also be the case that your use case requires an overall low latency. For example, when creating a resized version of a user’s profile image, you might not care if it takes 1 second or 5 seconds; on the other hand, when a user wants to check out a shopping basket, you don’t want to risk any delays as these might lead to abandonment and loss of revenue. Independent from the latency and determined by the workload is the access frequency. A certain concurrent requests, effectively establishing a permanent access pattern. Think of a user checking in at a certain location, triggering an update of a score, versus the case of an online chat environment.
To sum up the guidance that one can derive from the latency-versus-frequency graph, serverless computing is potentially a great fit for workloads that are in the lower-left quadrant of that is, use cases that are latency tolerant with a relatively low access frequency. The higher the access frequency and the higher the expectations around latency, the more it usually pays off to have a dedicated machine or container processing the requests. Granted, I don’t provide you with absolute numbers here, and the boundaries will likely be pushed in the future; however, this should provide you with a litmus test to check the general applicability of the paradigm. In addition, if you already have a serverless deployment, the infrastructure team might be able to supply you with data concerning the overall usage and costs. Equipped with this, you’ll be in a better position to decide if serverless computing continues to make sense from an economic point of view.
When (Not) to Go Serverless
There are a number of cases where serverless computing is a great fit, mainly centered around rather short-running, stateless jobs in an event-driven setup. These are usually found in mobile apps or IoT applications, such as a sensor updating its value once per day. The reason the paradigm works in this context is that you’re dealing with relatively simple operations executing for a short period of time. Let’s now have a look at some concrete application areas and use cases.
Application Areas and Use Cases
Typical application areas of serverless computing are: Infrastructure and glue tasks, such as reacting to an event triggered from cloud storage or a database Mobile and IoT apps to process events, such as user check-in or aggregation functions Image processing, for example to create preview versions of an image or extract key frames from a video Data processing, like simple extract, transform, load (ETL) pipelines to preprocess datasets
Let’s now have a closer look at a concrete example of how the paradigm is applied. LambCI is a serverless continuous integration (CI) system. , the creator of LambCI, was motivated to develop LambCI out of frustration with existing CI systems; in his own words:
You’ll be under- or overutilized, waiting for servers to free up or paying for server power you’re not using. And this, for me, is where the advantage of a serverless architecture really comes to light: 100% utilization, coupled with instant invocations. The architecture of LambCI is shown in it is essentially utilizing the Amazon Simple Notification Service (SNS) to listen to GitHub events and triggering a Lambda function that carries out the actual build, with the resulting build artifacts stored in S3 and build configuration and metadata kept in DynamoDB.
Figure 4-2. LambCI architecture
Limitations of LambCI at the moment are that there is no HTTP interface available (i.e., one has to interface with SNS), no root access can be provided (that is, it’s not suitable for building Docker images), and the build time is capped at five minutes. Nevertheless, since LambCI can be deployed based on a CloudFormation stack, using it can save a lot of money, especially for many shorter- running builds. Other exemplary use cases for serverless architectures include but are not limited to the following:
Forwardingo support chatops Blocking in CloudFlare Migrating anor small business Providing IRC notifications, as in
Carrying out Implementing a Realizing an IoT service, as in Doing Replacing Fetching nearbyta Integrating
Serverless computing is growing in popularity, and as we saw in
Chapter 2 , the number of offerings
is increasing. Does this mean that in the future we will eventually migrate everything to serverless? I don’t think so, and next we will have a look at challenges with the serverless paradigm that might help clarify why I don’t think this will be the case.
Challenges
While the serverless paradigm without doubt has its use cases and can help simplify certain workloads, there are naturally limitations and challenges. From most pressing to mildly annoying, these include:
Stateful services are best implemented outside of serverless functions. Integration points with other platform services such as databases, message queues, or storage are therefore extremely important. Long-running jobs (in the high minutes to hours range) are usually not a good fit; typically you’ll find timeouts in the (high) seconds range.
Logging and monitoring are a challenge: the current offerings provide little support for these operational necessities, and on top of that, the expectations are quite different than in traditional environments due to the short lifecycle. Local development can be challenging: usually developers need to develop and test within the online environment.
Language support is limited: most serverless offerings support only a handful of programming languages. positioning of the serverless approach (for example, on . This can serve as a baseline in terms of expectation management as well as a reminder of how young and fluent the ecosystem is.
Migration Guide
The process of migrating a monolithic application to a serverless architecture is by and large comparable with that of migrating to a , leaving stateful aspects aside.
event-driven, and batch-oriented nature of serverless functions. Furthermore, in comparison to breaking down a monolith into, say, 50 microservices, you might find yourself with hundreds of functions. In this situation, a migration of the whole system can be hard to manage and troubleshoot. A better approach might be to identify the workloads that are a good fit and migrate only this functionality.
Walkthrough Example
In this section, we will be using AWS Lambda for a simple walkthrough example to demonstrate the implications for operations, as outlined in
Chapter 3 . Note that the goal of the exercise is not to
provide you with an in-depth explanation of Lambda but to discuss typical workflows and potential challenges or limitations you might experience. The hope is that, equipped with this knowledge, you’ll be better prepared when you decide to apply the serverless paradigm in your own organization or project.
Preparation
For the walkthrough example, I’ll be using a blueprint: s3-get-object-python. This blueprint, as shown in , is written in Python and employs an S3 trigger to retrieve metadata for that S3 object when it is updated.
Figure 4-3. AWS Lambda dashboard: selecting a blueprint
Also, as a preparation step, I’ve created an S3 bucket called serops-we that we will be using shortly.Trigger Configuration
In the first step, depicted in , I configure and enable the trigger: every time a file is uploaded into the serops-we bucket, the trigger should fire. The necessary permissions for S3 to invoke the Lambda function are automatically added in this step.
Figure 4-4. Configuring the S3 trigger
Note that in this step I could also have applied certain filters, using the Prefix and Suffix fields, for example, to only react to events from a certain file type.
Function Definition
The next step, configuring the Lambda function, comprises a number of substeps, so let’s take these one by one. First we need to provide a name for the function (I’m using s3-upload-meta here; see Figure 4-5), and we can enter a description as well as selecting a runtime (Python 2.7 in our case).
Figure 4-5. Configuring the Lambda function: setting global properties
Next comes the actual definition of the function code, as shown in or the purpose of this example, I opted for the most primitive option, defining the code inline. Other options are to upload a ZIP file from local storage or S3. In a production setup, you’d likely have your CI/CD pipeline putting the code on S3.
can be arbitrarily chosen, the parameters are fixed in terms of order and type.
Figure 4-6. Providing the Lambda function code
Now we need to provide some wiring and access information. In this substep, depicted in , I declare the handler name as chosen in the previous step (lambda_handler) as well as the necessary access permissions. For that, I create a new alled lambda-we using a template that defines a read-only access policy on the S3 bucket serops-we I prepared earlier. This allows the Lambda function to access the specified S3 bucket.
Figure 4-7. Defining the entry point and access control
The last substep to configure the Lambda function is to (optionally) specify the runtime resource consumption behavior (see .
Figure 4-8. Setting the runtime resources
The main parameters here are the amount of available memory you want the function to consume and how long the function is allowed to execute. Both parameters influence the , and the (nonconfigurable) CPU share is determined by the amount of RAM you specify.
Review and Deploy It’s now time to review the setup and deploy the function, as shown in .
Figure 4-9. Reviewing and deploying the function The result of the previous steps is a deployed Lambda function like the one in
Figure 4-10. The deployed Lambda function Note the trigger, the S3 bucket serops-we, and the available tabs, such as Monitoring.
Invoke
Now we want to invoke our function, s3-upload-meta: for this we need to switch to the S3 service dashboard and upload a file to the S3 bucket serops-we, as depicted in
Figure 4-11. Triggering the Lambda function by uploading a file to S3
If we now take a look at the Monitoring tab back in the Lambda dashboard, we can see the function execution there ( . Also available from this tab is the “View logs in CloudWatch” link in the upper-right corner that takes you to the execution logs.
Figure 4-12. Monitoring the function execution Note that the logs are organized in so-called streams, and you can filter and search in them. This is especially relevant for troubleshooting.
Figure 4-13. Accessing the function execution logs
That’s it. A few steps and you have a function deployed and running. But is it really that easy? When applying the serverless paradigm to real-world setups within existing environments or trying to
teps
from the walkthrough example from an AppOps and infrastructure team perspective to make this a bit more explicit.
Where Does the Code Come From?
At some point you’ll have to specify the source code for the function. No matter what interface you’re using to provision the code, be it the command-line interface or, as in , a graphical user interface, the code comes from somewhere. Ideally this is a (distributed) version control system such as Git and the process to upload the function code is automated through a CI/CD pipeline such as
.
and
when, and you can roll back to a previous version if you experience troubles with a newer version.
Figure 4-14. Automated deployment of Lambdas using Jenkins (kudos to AWS)
How Is Testing Performed?